DNA typically comprises both methylated and unmethylated bases. Prokaryotic DNA is methylated at cytosine and adenosine residues (see, e.g., McClelland et al., Nuc. Acids. Res. 22:3640-3659 (1994). Methylation of prokaryotic DNA protects the DNA from digestion by cognate restriction enzymes, i.e., foreign DNAs (which are not methylated in this manner) that are introduced into the cell are degraded by restriction enzymes which cannot degrade the methylated prokaryotic DNA. DNA methylation patterns can be used to identify specific bacterial types (e.g., genus, species, strains, and isolates).
Mammalian DNA can only be methylated at cytosine residues, typically these cytosines are 5′ neighbors of guanine (CpG). This methylation has been shown by several lines of evidence to play a role in gene activity, cell differentiation, tumorigenesis, X-chromosome inactivation, genomic imprinting and other major biological processes (Razin and Riggs eds. in DNA Methylation Biochemistry and Biological Significance, Springer-Verlag, N.Y., 1984).
In eukaryotic cells, methylation of cytosine residues that are immediately 5′ to a guanosine, occurs predominantly in CG poor loci (Bird, Nature 321:209 (1986)). In contrast, discrete regions of CG dinucleotides called CpG islands remain unmethylated in normal cells, except during X-chromosome inactivation and parental specific imprinting (Li, et al., Nature 366:362 (1993)) where methylation of 5′ regulatory regions can lead to transcriptional repression.
Aberrant methylation, including aberrant methylation at specific loci, is often associated with a disease state. For example, de novo methylation of the Rb gene has been demonstrated in a small fraction of retinoblastomas (Sakai, et al., Am. J. Hum. Genet., 48:880 (1991)), and a more detailed analysis of the VHL gene showed aberrant methylation in a subset of sporadic renal cell carcinomas (Herman, et al., PNAS USA, 91:9700 (1994)). Expression of a tumor suppressor gene can also be abolished by de novo DNA methylation of a normally unmethylated 5′ CpG island. See, e.g., Issa, et al., Nature Genet. 7:536 (1994); Merlo, et al., Nature Med. 1:686 (1995); Herman, et al., Cancer Res., 56:722 (1996); Graff, et al., Cancer Res., 55:5195 (1995); Herman, et al., Cancer Res. 55:4525 (1995). Methylation of the p16 locus is associated with pancreatic cancer. See, e.g., Schutte et al., Cancer Res. 57:3126-3131 (1997). Methylation changes at the insulin-like growth factor II/H19 locus in kidney are associated with Wilms tumorigenesis. See, e.g., Okamoto et al., PNAS USA 94:5367-5371 (1997). The association of alteration of methylation in the p15, E-cadherin and von Hippel-Lindau loci are also associated with cancers. See, e.g., Herman et al., PNAS USA 93:9821-9826 (1997). The methylation state of GSTP1 is associated with prostate cancer. See, e.g., U.S. Pat. No. 5,552,277. Tumors where certain genomic loci are methylated have been found to respond differently to therapies such as cis-platin or radiation treatment than tumors where the same genomic loci are unmethylated. It is clear that DNA from tumor cells at certain genomic loci can be different in the levels of DNA methylation and in this way can be distinguished from the DNA from adjacent normal cells. DNA from tumor cells has been found in various body fluids and other clinical specimens collected from cancer patients. For example, methylated DNA having the same sequence of tumor suppressor genes has been found in serum, urine, saliva, sputum, semen, lavages, cell scrapes, biopsies, resected tissues, and feces. Therefore, detection of altered methylation profiles at loci where such alterations are associated with disease can be used to provide diagnoses or prognoses of disease.
Current methods for determining whether DNA is methylated or unmethylated typically use methylation-sensitive restriction enzymes or a combination of methylation-sensitive and methylation-insensitive restriction enzymes (see, e.g., Burman et al., Am. J. Hum. Genet. 65:1375-1386(1999); Toyota et al., Cancer Res. 59:2307-2312 (1999); Frigola et al., Nucleic Acids Res. 30(7):e28 (2002); Steigerwald et al., Nucleic Acids Res. 18(6):1435-1439 (1990); WO 03/038120; and U.S. patent Publication No. 2003/0129602 A1). Methylation-sensitive restriction enzymes cleave their cognate DNA recognition sites only if specific nucleotides within those sites are not methylated. Therefore, methods used to detect the presence of DNA methylation following methylation-sensitive restriction enzyme digestion rely on reporting a negative enzymatic outcome. That is, methylation is detected based on the failure of the methylation-sensitive restriction enzyme to cleave its DNA recognition sequence. This strategy introduces the unavoidable caveats of basing a positive experimental measurement on a negative enzymatic outcome (i.e. the result that reports the presence of DNA methylation is equivalent to the result that would occur if the enzyme was absent or inactive due to suboptimal conditions).
In some cases, methylation-sensitive restriction enzymes are used in combination with methylation-insensitive restriction enzymes. Methylation-insensitive restriction enzymes cleave their DNA recognition sites regardless of the presence of DNA methylation. Combining digestion by a methylation-sensitive restriction enzyme with digestion with a methylation-insensitive restriction enzyme that cleaves the same DNA recognition site (an isoschizomer) allows confirmation that the DNA site of interest is susceptible to restriction enzyme digestion in general, but does not alleviate the caveats associated with use of methylation-sensitive enzymes as the sole indicator of the presence of DNA methylation. In addition, these methods act on non-randomly fragmented DNA and can not measure DNA methylation of sequences in much of the genome.
Thus, there is a need in the art for more efficient and more comprehensive methods of detecting methylation of DNA, particularly DNA at specific loci. The present invention addresses these and other needs.